Many patients relapse over time despite initial response to systemic therapy. One explanation is that a rare sub-population of cancer stem cells with tumorigenic potential is intrinsically resistant to therapy. Consistent with this, we have shown for the first time clinically in human breast cancer patients that residual tumors after chemotherapy are 1) enriched for the tumorigenic CD44+/CD24-/low population, 2) show enhanced mammosphere-forming efficiency (MSFE), and 3) display an increase in outgrowths in xenograft transplants in immunocompromised SCID/Beige mice, thus suggesting their increased tumorigenicity. Our recent data from paired human breast cancer samples indicates that standard therapy eliminates dividing daughter cells, so that samples obtained after therapy are enriched for CD44+/CD24-/low putative "breast cancer stem cells" that have the ability to self-renew in mammosphere cultures, and to give rise to tumors upon xenograft transplantation. We have identified a cancer stem cell signature of CD44+/CD24-/low mammosphere-forming cells derived from human breast cancer biopsies. The top canonical pathways identified include Notch and PI3-AKT, and other signaling pathways. We now propose a series of preclinical and clinical studies to directly test the hypothesis that breast cancer stem cells can be specifically targeted by inhibitors of the Notch, PI3-AKT and other pathways. A unique component of our studies is the availability of human biopsy samples obtained before and after targeted therapy in breast cancer patients with residual disease after preoperative (neoadjuvant) therapy. Specific Aims and Study Design 1. To determine whether suppression of identified self-renewal and treatment resistance pathways can improve existing cancer therapies in preclinical models. Beginning with Notch and PI3-AKT inhibitors, we will determine if these will improve efficacy of conventional therapy, using MSFE and human breast cancer xenograft models. Next, we will target the top ~300 genes differentially expressed in our stem cell signature by ordered lentivirus-based shRNA libraries designed to allow genetic "knockdown" of every gene in the human genome, as well as the development of high-throughput functional genomic assays of stem cell self-renewal. 2. To conduct novel clinical trials to determine whether suppressing stem cell self-renewal and treatment resistance pathways can improve existing cancer therapies in breast cancer patients. Clinical trials with novel inhibitors of stem cell self-renewal (Notch and PI3) have been planned. We propose to include patients with advanced breast cancers refractory to conventional therapy, as these women have a poor expected clinical outcome, and who are most likely to benefit from therapies targeting self-renewal pathways. 3. To perform correlative studies using breast cancer biopsy specimens from these clinical trials. Correlative studies using human cancer biopsies from these trials will be conducted, e.g. a decrease in stem cell markers and tumorigenic potential, as well as downstream effects of inhibition of the relevant pathways.